G, what effect would this have upon the ultimate

Transcription

1 STUDIES ON BACTERIA DEVELOPING RESISTANCE TO PENI- CILLIN FRACTIONS X AND G IN VITRO AND IN PATIENTS UNDER TREATMENT FOR BACTERIAL ENDOCARDITIS By HARRY F. DOWLING, HAROLD L. HIRSH, AND C. BARBARA O'NEIL (From the George Washington Medical Division, Gallinger Municipal Hospital and the Department of Medicine, George Washington University, Washington) Although penicillin G is the fraction of penicillin which is best known and most used, 2 other known fractions of this antibiotic have been isolated in crystalline form, penicillin F and penicillin X. Since the latter fraction is now being manufactured in appreciable quantities, it is important to determine whether it is superior in any way to penicillin G. We have attacked this problem by attempting to answer the following questions: (1) What are the relative sensitivities of different bacteria to penicillins X and G? (2) If the resistances of bacteria are raised in titro by the use of penicillin X or penicillin G, what effect would this have upon the ultimate sensitivities of the organisms to the 2 fractions? (3) If the resistance of an organism should be raised during the course of treatment of a patient with 1 of these fractions of penicillin, would there be any change in the relative sensitivity of the organism to the 2 fractions? (4) Is there any difference in the absorption, distribution and excretion of penicillin X as compared with penicillin G? (5) Is there any difference in the clinical results obtained with the use of the 2 fractions? The last 2 questions have been discussed in another publication (1). In the present paper we shall give the results which we have obtained in seeking answers to the first 3 questions. MATERIALS AND METHODS Several preparations of penicillin X were used. One commercial preparation contained 75 per cent of sodium penicillin X; others, 90 and 95 per cent.1 Several brands of commercial sodium penicillin G, made by various manufacturers, were used. Crystalline calcium penicillin G, obtained from the Food and Drug Administration, was also employed. In a few of the tests, a preparation of 1 Supplied by the Lederle Laboratories, Inc. (Received for publication January 25, 1946) crystalline sodium penicillin G2 was compared with the crystalline calcium penicillin G, and found to give identical results. The tests for penicillin sensitivity combined in each tube 2 reagents: (a) 0.2 from a series of 2-fold dilutions of penicillin, and (b) 0.5 ml. of a standard dilution of each micro-organism in a medium containing 0.2 per cent of horse blood. A 24-hour culture of staphylococci (or other bacteria which grew readily) in tryptose phosphate blood broth was diluted in the same medium to the 10' dilution; streptococci, to 10'; pneumococci and meningococci, to 10; influenza bacilli, to 10'. To the final dilution to be used, 0.2 per cent of the horse blood was added, and 0.5 ml. aliquots transferred to tubes already containing the required dilutions of penicillin. After 18 hours incubation the end-point tube showing no hemolysis or turbidity was recorded as containing the minimal effective concentration of penicillin. The end-points were checked by streaking the last negative tube on a bloodagar plate. To increase the resistance of the bacterial strains to penicillins X and G, each was first tested for sensitivity and then transferred serially through: (1) a tube of an appropriate medium which contained the greatest amount of penicillin X which would allow growth; (2) a tube of same medium containing the greatest amount of penicillin G which would allow growth; and (3) a control tube of the medium containing no penicillin. The media employed were tryptose phosphate broth in the case of staphylococci, brain-heart infusion broth for streptococci and meningococci, and the same medium plus 1 per cent agar for pneumococci. At 24-hour intervals (or, in the case of meningococci, at 48-hour intervals), 02 ml. of the control culture was transferred to another tube of control me- 665 dium. If the organisms in the tubes containing penicillin had grown well, 0.2 ml. of the culture was transferred from each tube to another tube which contained double the concentration of the same fraction of penicillin. If they had not grown well, the transfer was made to a tube containing the same concentration as on the previous day. This procedure was continued until the organisms persisted in growing poorly when transferred in the same concentration of penicillin for 3 successive days. When this plateau had been reached in each of the fractions of penicillin, the experiment was terminated as far as that particular strain was concerned, and the sensitivities of 2 Supplied by Merck and Company

2 666 HARRY F. DOWLING, HAROLD L. HIRSH, AND C. BARBARA O'NEIL the organisms which had been grown in the 2 fractions of penicillin, and of the control organisms, were tested. The bacteria used in determining sensitivities and for the elevation of resistance were obtained from patients with various kinds of infections, with the exception of a few of the staphylococci and non-hemolytic streptococci. These were cultured from the blood of healthy persons with a transient bacteremia following the extraction of teeth. RESULTS OF STUDIES ON COMPARATIVE RESISTANCE Several investigations have been reported in which the comparative resistance of certain organisms to penicillins X and G were determined-. Welch and his associates (2) stated that the X fraction was more effective than commercial penicillin against 1 strain each of Klebsiella pneumoniae, Bacillus cereus and pneumococcus Type I. Libby and Holmberg (3) tested 1 strain each of 10 different bacteria, and found only slight differences in their relative sensitivities to the 2 fractions. Ory and his co-workers (4) studied 209 strains, and found that the streptococci, pneumococci, gonococci and meningococci were generally from 2 to 8 times as sensitive to penicillin X as to penicillin G, on a unit-for-unit basis, while the staphylococci, 3 strains of Friedlander's bacillus, and 1 of Hemophilus influenzae, showed essentially the same resistance to the 2 fractions. We have placed our results in the same form (Table I) as was used by Ory, in order to make comparison easy. One hundred and one strains were tested for resistance to penicillins X and G. Commercial penicillin preparations were used in the case of 56; crystalline penicillin preparations, in the case of 14; and in the case of 31 strains all 4 of these preparations were employed. Inasmuch as no significant difference could be observed in the results obtained with the crystalline and commercial preparations of either fraction, no distinction has been made between them in the table. When a comparison of sensitivity to penicillin X and penicillin G is made on the basis of units, most of the organisms were more sensitive to penicillin X. Only 4 strains were more sensitive to the G fraction: 1 strain each of staphylococcus and Hemophilus influenzae and 2 strains of Streptococcus viridans. Seventeen strains (including the single strain of C. diphtheriae tested) were equally sensitive to the 2 penicillin fractions, while 80 strains were from 2 to 32 times more sensitive to penicillin X than to penicillin G. The antibacterial effects of the 2 fractions may be better compared, however, on the basis of weight. Each mgm. of crystalline penicillin G has a potency of about 1,650 units; each mgm. of crystalline penicillin X, approximately 900 units (2)..BLE I Comparative sensitivity of various bacteria to penicillin X and penicillin G Beta- Sensitivity, Staphy- h~em~o- Streplo- Pneumo- Gono- Menin- E. H. C. All Sensitivity, arranged according lococcus Jytic COCCUS coccus coccus gococcus Syphi influ- diph- bac- arranged according to units strepto- viridans ensae theriae teria to weight coccus More sensitive to X 32-fold fold 16-fold fold 8-fold fold 4-fold fold 2-fold Same sensitivity -I --I 1 1--I 1 -- I It G I Same sensitivity ! toxand G 17: 61I Ill1 More sensitive to G 2-fold

3 TABLE II Range of sensitivity of 101 strains of bacteria to penicillin X and penicillin G Organism BACTERIA DEVELOPING RESISTANCE TO PENICILLIN Amount of penicillin required to inhibit growth of least susceptible strain and most susceptible strain Penicillin X units per ml. Penicillin G units per ml. Staphylococcus 3.6 to to 0.01 Beta-hemolytic streptococcus Group A 0.09 to to Streptococcus viridans 0.7 to to Pneumococcus Types I, III, IV, VII, and XIV 0.1 to to Gonococcus to to Meningococcus Groups I II and II alpha 0.04 to to 0.01 E. typhi 1.4 to to 2.2 H. influenzae Type B 1.1 to to 0.6 C. diphtheriae (one strain tested) Accordingly, a given weight of penicillin G is approximately twice as potent in terms of units as the same weight of penicillin X. In the extreme right-hand column of Table I we have arranged the legend so as to compare the relative sensitivities of the strains according to the weights of the 2 fractions used in the tests. Measured in this way, the 2 fractions showed less difference in TABLE III Development of resistance to penicillin X and G in vitro Source Number Organism Organism transfers made Material Disease their effectiveness upon the various bacteria. Twenty-five per cent of the strains showed equal sensitiveness to the 2 fractions, 54 per cent of the strains were from 2 to 16 times as sensitive to X as to G, and 21 per cent were from 2 to 4 times as sensitive to G as to X. The range of the sensitivities of the various organisms to the 2 penicillin fractions is shown in Table II. All strains of beta hemolytic streptococci, gonococci and meningococci were very sensitive to penicillin X, and usually to penicillin G also. Staphylococcus and streptococcus strains varied from resistant to very sensitive. The few strains of Hemophilus influenzae and Eberthella typhi investigated were moderately resistant to penicillin X, and even more resistant to the G fraction. The 1 strain of C. diphtheriae encountered was moderately sensitive to both fractions. RESULTS OF RAISING THE RESISTANCE OF BACTERIA in vitro Ten strains of staphylococci, 4 of pneumococci, 2 of meningococci and 1 each of alpha-hemolytic and beta-hemolytic streptococci were made more resistant to both fractions of penicillin. Table III shows the original sensitivities of these Original Sensitivity after transfers sensitivity In penicillin In penicillin In control XG x G mdu medium~~~~~~~~o of Sensitivity tested by using penicillin: X G X G X G X G units per ml. units per ml. units per ml. units per ml. Staphylococcus Pus Abscess of leg Staphylococcus Blood Transient bacteremia following tooth extraction Staphylococcus Blood Transient bacteremia following tooth extraction Staphylococcus Pus Abscem of skin Staphylococcus Blood Endocarditis Staphylococcus Urine Pyelitis Staphylococcus Blood Transient bacteremia following tooth extraction Staphylococcus Throat Acute pharyngitis Staphylococcus Throat Acute pharyngitis Staphylococcus Sputum Bronchiectasis Pneumococcus Type I Blood Pneumonia Pneumococcus Type III Spinal fluid Meningitis Pneumococcus Type XII Spinal fluid Meningitis Pneumococcus Type XII Blood Pneumonia Beta-hemolytic streptococcus, Group A, Type II Throat Scarlet fever Alpha-hemolytic streptococcus Urine Pyelitis Meningococcus Group I Spinal fluid Meningitis Not done Meningococcus Group II Spinal fluid Meningitis

4 668 HARRY F. DOWLING., HAROLD L. HIRSH, AND C. BARBARA 0'EI :2 0 Oe4ffl4': :- 42I a I-I--i-I~~~~~~~~~~~~~~c Ii a ~~~~~~~6. c : A~~~I PC*

5 BACTERIA DEVELOPING RESISTANCE TO PENICILLIX strains to penicillins X and G, and the sensitivities hand column in Table IV) it is seen that the number of strains possessing greater sensitivity to after they had been made resistant. The control cultures which were passed along in broth without penicillin often showed changes in sensitivity approximately the same as the number of those penicillin X before their resistance was raised, was to penicillin when tested at the end of the experiment. Usually these were slight, there being a After their resistance had been raised to penicillin possessing greater sensitivity to penicillin G. 2- or 4-fold increase or decrease, as compared with X, 1 was more sensitive to penicillin X, 13 were the original sensitivity. In the case of 3 of the more sensitive to penicillin G, and 3 were equally staphylococcus strains, the sensitivity of the control cultures increased 4- to 18-fold. No explanasistance had been raised to penicillin G, 5 were sensitive to the 2 fractions; whereas after their retion can be given for this phenomenon. In every more sensitive to penicillin X, 9 to penicillin G and instance, except 1, the resistance of the strains 4 were equally sensitive to the 2 fractions. It is transferred in penicillin was raised above that of evident from these figures that raising the resistance of these strains to penicillin X caused them the control cultures. The degree of resistance which could be induced to be relatively more sensitive to penicillin G than varied greatly from strain to strain and also in the to penicillin X, whereas raising their resistance to same strain, depending upon which fraction was penicillin G left the proportions of those resistant used. The resistance of some strains was raised to the 2 fractions about the same as before. higher by means of penicillin X, and that of others was raised higher with the use of penicillin G. Studies on strains which became resistant in patients under treatment with penicillin When the resistance of a strain had been raised by employing 1 fraction, there was little difference While we were investigating the relative sensitivity of bacteria to penicillin X and penicillin G, in the sensitivity of that strain to the 2 different fractions. For instance, the sensitivity of 1 staphylococcus strain was 28.6 units per ml. of either endocarditis caused by a Streptococcus viridans we encountered a patient, J. F., with a bacterial penicillin X or penicillin G when its resistance had which developed increasing resistance to penicillin while the patient was under treatment with been raised by means of penicillin X, and 3.6 units per ml. of each penicillin fraction when its resistance had been raised by the employment of penicil-~ this antibiotic. lin G. This patient was a 22-year old colored male with a In Table IV we have compared the relative sensitivity of the organisms to the 2 fractions of penimatic fever. Three weeks before admission he developed questionable history of syphilis, and no history of rheucillin before and after the induction of resistance. headache, followed by pain and stiffness in the cervical spine and swelling of both feet and ankles. He was acutely When the comparison is made on a unit basis, 11 ill on admission to the hospital, with a temperature of these strains were more sensitive to penicillin X of 102 F, a pulse rate of 112, and a blood pressure of 140 than to penicillin G before their resistance was systolic and 60 diastolic. There was a systolic thrill in raised, and 7 were equally sensitive to the 2 fractions. After the resistance of these strains was systolic and diastolic murmurs, maximal in the same area the 3rd and 4th interspaces to the left of the sternum, and and transmitted over the entire precordium. The liver raised to penicillin X, only 4 were more sensitive edge was 4 cm. below the right costal margin. The spleen to that fraction than to penicillin G, while 10 were was'not felt. Roentgenogram of the chest revealed slight equally sensitive to the 2 fractions, and 3 were enlargement of the left ventricle and of the pulmonary more sensitive to penicillin G. On the other hand, conus area, and a moderate increase in the pulmonary after the resistance of the strains had been raised vascular shadows. When 6 blood cultures, taken on different days, had to penicillin G, 9 were more sensitive to penicillin grown the Streptococcus viridans, as shown in Figure 1, X, 8 were equally resistant to the 2 fractions, and 1 the patient was started on penicillin G by mouth. For the was more sensitive to G than to X. first 24 hours the dose was 100,000 units in amphojel every As has been stated before, the weight of each 4 hours; and for the next 24 hours, 200,000 units every fraction of penicillin used in testing the strains is 4 hours. Since this did not seem to be influencing the course of the disease, the dose was changed to 25,000 a truer method of comparison. When the comparison is made in this way (the extreme right- test of the etiologic organism at this time for units of penicillin G intramuscularly every 2 hours. A sensitivity 660

6 670 II~~ARRY D~OWLING, HARZOLfl L. IIIRSH, AND C. BARBARA O'NEIL F.3.T C. M.22 BAC.TE.RIAL ENDOCARDITIS WEEK 1 St 2,nj I PA 4t6 St6 Gi6 7t6 4B t6 9t6 loth 1o4- IL L I 0ad c 0 culture Fl P Sensitivity.to PX to P c, o 14-0 I-IA WEEK 11 th 12 L6 13kh 14th 15% IG th th 10A I'm I's 11% r II.I I L 0 0 culture iq IPi I I I 0111 I I III i-o - 0 io O LLJt 4- FIGURE 1. to penicillin revealed that unit per ml. of either the X or G fraction was sufficient to inhibit its growth. The temperature fell and the blood cultures became negative on the intermittent intramuscular injections, but after a week fever and bacteremia reappeared. The sensitivity of the organism cultured at this time had risen to 0.01 unit per ml. in the case of penicillin X, and 0.04 unit in the case of penicillin G. Penicillin X was tried in doses of 500,000 units per day by continuous intramuscular infusion without success. Positive blood cultures continued to make their appearance, and the sensitivity of the organisms reached 0.04 unit per ml. and 0.18 unit per ml. for the X and G fractions, respectively. Due to the scarcity of penicillin X at that time, treatmnent was resumed with penicillin G in doses of 500,000 units per day by continuous intramuscular infusion. The sensitivity of the organism was again tested and f'ound to be 1.4 units per ml. for both the X and G fractions. As a result of this, the dose was doubled. After a few days on 1,000,000 units per day by -intramuscular infusion, the patient itn-

7 TABLE V Relative sensitivities to penicilin X and penicillin G of a strain of Staphylococcus aureus isolated from a patient with endocarditis Sensitivity of organism BACTERIA DEVELOPING RESISTANCE TO PENICILLIN Day of Sensitivity to Sensitivity to disease penicillin X penicillin G units per ml. units per ml proved considerably, the temperature fell, and the blood culture became and remained sterile. After 9 weeks of this regime, the patient left the hospital against advice. He was examined 4 months later and found to be free of symptoms and of all evidence of infection. With the exception of the first few days, when the patient was being treated orally, the concentrations of penicillin in the blood were always higher than the amount of penicillin required to inhibit the growth of the organism it; vitro. In spite of this, the infection remained active until a dose of 1,000,000 units per day was given. Another patient, L. M., a 34-year old colored female, had an endocarditis caused by Staphylococcus aureus, an organism which became more resistant to penicillin during the course of treatment. When this organism was first cultured from the blood, on the 15th day of the disease, it was sensitive to 0.01 unit per ml. of penicillin X and to 0.04 unit per ml. of penicillin G (Table V). On the 29th day, its resistance to each fraction had increased 2-fold. Organisms cultured from the blood at intervals during the next 12 days showed gradually increasing resistance to both the X and G fractions until 0.2 and 0.4 units per ml. respectively, were required to inhibit the growth of the organisms. During the period in which these cultures were obtained, the patient was being treated with increasingly larger doses of penicillin G, starting with 15,000 units every 3 hours by intramuscular injection on the 15th day of the disease. Before the organisms were cultured from the blood for the last time (on the 41st day), the patient had been receiving 1,000,000 units per day by continuous intramuscular infusion. She was continued on penicillin treatment for 8 more weeks, during which she became afebrile and symptom-free, and her blood cultures remained negative. She has recently been discharged from the hospital, apparently infection-free. DISCUSSION The excellent results obtained with the commonly used G fraction of penicillin do not preclude the possibility that some other fraction may be still more effective. It is necessary, therefore, to determine whether bacteria causing human infections are more sensitive to the X fraction than to penicillin G, or whether they develop resistance against 1 fraction without developing it against the other. Upon comparing the relative sensitivity of various bacteria from human sources to approximately equal amounts of penicillin X and penicillin G by weight, we found that slightly more than '/2 of the strains were more sensitive to X than to G, /4 were equally sensitive to both fractions, and slightly less than ¼ were more sensitive to G than to X. When our figures are compared with Ory's (4), using the weight of the penicillin fraction as a basis of measurement in both instances, the results are very similar. Libby and Holmberg (3) compared the sensitivity of a few organisms to known weights of penicillin X and G, and found of penicillin G to the number of micrograms of penicillin X required to inhibit growth of the organisms varied from 0.6 to 2.0. It is evident from these 3 investigations that, while some strains of bacteria tested showed considerably more sensitivity to penicillin X than to penicillin G, and some showed more sensitivity to G than to X, nevertheless most that the ratio of the number of jug. 671 strains showed very little difference in sensitivity to the 2 fractions. Since most of the bacteria which are causing human infections at the present time have presumably not yet come in contact with penicillin, it is possible that great differences in their sensitivities to the various fractions of penicillin will develop only after prolonged contact with 1 of these fractions. We might expect the organism to become relatively more resistant to the fraction with which it had been in contact. Actually, as shown in Table III, when a strain's resistance to 1 fraction was increased, its resistance to the other fraction increased also. Further evidence is provided by the cases of the 2 patients with bacterial endocarditis: the organisms developed resistance to both penicillin X and G while the patients were being treated with the G fraction. Flippin and his associates (5) reported the development of resistance to penicillin G by the organism of a patient under treatment with this fraction. Before treatment, unit of penicillin G per ml. was sufficient to inhibit its growth. Six months later 0.75 unit per ml. was required. At the same time, however, growth of the organism was inhibited by

8 672 HARRY F. DOWLING, HAROID L. HIRSH, AND C. BARBARA O'NEIL only 0.05 unit per ml. of penicillin X. Treatment was changed from penicillin G to penicillin X, and soon after this the patient improved and apparently recovered completely. Since no determination of sensitivity to penicillin X was made on the organism recovered before penicillin was administered, it is impossible to tell whether the organism developed any resistance to this fraction during the course of treatment. In the case of our patient, with Streptococcus siridans endocarditis, the organism was found to develop resistance to the 2 fractions simultaneously, the sensitivity finally reaching 1.43 units per ml. of either fraction. Fortunately, the patient recovered after prolonged treatment with 1,000,000 units of penicillin G per day. The organism in the case of our patient with Staphylococcus aureus endocarditis also developed increasingly greater resistance to both the X and G fractions simultaneously. It has already been pointed out that, following passage of the organisms through penicillin X, resistance to penicillin X was developed to a measurably greater degree than resistance to penicillin G. Similar results were obtained when strains were passed through penicillin G. In every, instance, however, there was a simultaneous increase in resistance to both fractions, and this increase was greater than the difference between the fractions. SUMMARY AND CONCLUSIONS 1. One hundred and one strains of bacteria from human sources were tested for sensitivity to the 2 penicillin fractions X and G. When the weights of the penicillin fractions employed were used as the basis for comparison, 55 strains were from 2 to 16 times more sensitive to penicillin X than to penicillin G, 21 were 2 to 4 times more sensitive to penicillin G, and 25 were equally sensitive to each fraction. 2. Sixteen strains were made resistant to penicillin X and penicillin G by serial passage in media containing increasingly larger amounts of these fractions. At the termination of the experiments, some of the strains whose resistance had been raised to penicillin X had become relatively more sensitive to penicillin G than to penicillin X. Among the strains made resistant to penicillin G, there was no significant difference between the number which were more sensitive to penicillin X than to penicillin G at the beginning of the experiments, and the number which were more sensitive to penicillin X at the end of the experiments. In every instance when the resistance to 1 fraction was raised, the resistance to the other fraction followed along, to the same or nearly the same extent. 3. The sensitivity of the etiologic organisms to penicillins X and G were studied in a patient with Streptococcus viridans endocarditis and in a patient with Staphylococcus aureus endocarditis. In each instance, while the patient was receiving penicillin, the organism responsible for the infection developed resistance to both penicillin fractions simultaneously. 4. It is concluded that organisms which cause human infections usually do not show great differences in their relative sensitivity to penicillin G and penicillin X. In most instances, when resistance to 1 of these fractions increases, resistance to the other fraction will also increase. We wish to thank the members of the Georgetown Medical Division for the opportunity of studying one of the cases of endocarditis. BIBLIOGRAPHY 1. Hirsh, H. L., Dowling, H. F., and Sweet, L. K., Treatment of various infections with penicillin X, with a preliminary note on the value of penicillin X in scarlet fever. Ann. Int. Med., 1946, 25, Welch, H., Putnam, L. E., Randall, W. A., and Herwick, R. P., Penicillin X. Successful treatment of gonorrhea with a single intramuscular injection. J. A. M. A., 1944, 126, Libby, R. L., and Holmberg, N. L., The activity of penicillins G and X in vitro. Science, 1945, 102, Ory, E. M., Meads, M., and Finland, M., Penicillin X. Comparison with penicillin G with respect to sensitivity of pathogenic organisms and serum levels. J. A. M. A., 1945, 129, Flippin, H. F., Mayock, R. L., Murphy, F. D., and Wolferth, C. C., Penicillin in the treatment of subacute bacterial endocarditis. A preliminary report on twenty cases treated over one year ago. J. A. M. A., 1945, 129, 841.